Three researchers at Stockholm University and KTH Royal Institute of Technology participated in the talk on ”The promise and threat of artificial intelligence” together with a Swiss Professor joining via Skype. The discussion highlighted the increasing development of artificial intelligence technologies, focusing on the positive and interesting applications of AI, as well as on some of the negative aspects.

“Today, we are able to predict on a global scale what we are going to say. Systems are already predicting who I should be in contact with on LinkedIn, what music I would like on Spotify. This intelligence taps on what my wishes are, and I am not in control anymore. We want service so we are handing over our data, we surrender to technology" says Theo Kanter, Professor of Computer Science at the Department of Computer and Systems Sciences Stockholm University, describing the modern use of Artificial Intelligence.

Not a threat to humanity

According to Jürgen Schmidhuber, Professor in Artificial Intelligence, Scientific director at the Swiss AI Lab IDSIA, and Skype guest of the event, there is no reason to see AI as a grave threat to human beings: "I have little doubt that in a few decades, humans will not be the main decision makers anymore. Time will not stop, of course, but there will be some sort of rupture in human history, which has been dominated by humans for at least 40 000 years".

Artificial intelligence is a shift in our world understanding. It can be defined as behavior embodied by machines and software, robots adapting to the context.

Human robots in domestic areas

Kristina Nilsson Björkenstam, PhD in Computational Linguistics Stockholm University says: “It would be very interesting to put a baby robot into my home”. Indeed, scientists want to integrate robots into domestic environments and make them act like humans: "Learning a language has a lot to do with interpreting what the counterpart would want you to do. One problem with learning a language is that words have many meanings. There are many possible analysis of one sentence, but human can understand the meaning by excluding the wrong interpretations. It would be interesting if we could get robots to interpret different meaning”.

"Robots have problems understanding its envinronment and what it needs to do. I can programme a robot to do certain things, but we cannot get artificial intelligence to understand what they can do with a bottle for instance", explained Christian Smith, Assistant professor in Computer Science KTH Royal Institute of Technology.

Improving living conditions

Commenting on the negative perspectives of the concept of artificial intelligence, Christian Smith concluded: "In any kind of development and progress in technology, people has been worried of the next step, but no one has been willing to let go of the last step".

The development of artificial intelligence technologies can certainly be used to provide humanity with better living conditions. Artificial intelligence and its changes within the society shouldn’t be considered as a threat to human beings but as a tool to support the planet's development.

All Crosstalks episodes are available here.